Advanced analysis and event reconstruction for the CTA Observatory

TL;DR

This paper discusses advanced data analysis techniques for the CTA gamma-ray observatory, improving event reconstruction and gamma/hadron separation to enhance sensitivity over current methods.

Contribution

It introduces the Paris-MVA analysis method, combining Hillas parameters, 3D-Model algorithms, and Boosted Decision Trees for improved performance in CTA data analysis.

Findings

Sensitivity improved by a factor of about 2

Effective gamma/hadron separation achieved

Algorithms successfully applied to H.E.S.S. data

Abstract

The planned Cherenkov Telescope Array (CTA) is a future observatory for very-high-energy (VHE) gamma-ray astronomy composed of one site per hemisphere. It aims at 10 times better sensitivity, a better angular resolution and wider energy coverage than current installations such as H.E.S.S., MAGIC and VERITAS. In order to achieve this level of performance, both the design of the telescopes and the analysis algorithms are being studied and optimized within the CTA Monte-Carlo working group. Here, we present ongoing work on the data analysis for both the event reconstruction (energy, direction) and gamma/hadron separation, carried out within the HAP (H.E.S.S. Analysis Package) software framework of the H.E.S.S. collaboration, for this initial study. The event reconstruction uses both Hillas-parameter-based algorithms and an improved version of the 3D-Model algorithm. For the gamma/hadron discrimination, original and robust discriminant variables are used and treated with Boosted Decision Trees (BDTs) in the TMVA (Toolkit for Multivariate Data Analysis) framework. With this advanced analysis, known as Paris-MVA, the sensitivity is improved by a factor of about 2 in the core range of CTA relative to the standard analyses. Here we present the algorithms used for the reconstruction and discrimination, together with the resulting performance characteristics, with good confidence, since the method has been successfully applied for H.E.S.S.